[project @ 2001-10-17 13:13:36 by simonmar]

Add a comment about assumptions that we make about the result of
etaExpand.

diff --git a/ghc/compiler/coreSyn/CorePrep.lhs b/ghc/compiler/coreSyn/CorePrep.lhs
index 5a4b636..298e599 100644 (file)
--- a/ghc/compiler/coreSyn/CorePrep.lhs
+++ b/ghc/compiler/coreSyn/CorePrep.lhs
@@ -427,6 +427,12 @@ mkLocalNonRec bndr dem floats rhs
        --     code compiled without -O.  The case in point is data constructor
        --     wrappers.
        --
+       -- NB2: we have to be careful that the result of etaExpand doesn't
+       --    invalidate any of the assumptions that CorePrep is attempting
+       --    to establish.  One possible cause is eta expanding inside of
+       --    an SCC note - we're now careful in etaExpand to make sure the
+       --    SCC is pushed inside any new lambdas that are generated.
+       --
     getUniquesUs               `thenUs` \ us ->
     let
        rhs' = etaExpand (exprArity rhs) us rhs bndr_ty